复杂纹理环境的光流/惯性组合导航方法

为解决无人机(unmanned aerial vehicle,UAV)在全球定位系统(global positioning system,GPS)信号弱或拒止环境下实现自主导航的问题,提出一种适用于复杂纹理环境的光流/惯性组合导航方法.通过基于低纹理场景的改进卢卡斯-卡纳德(lucas and kanade,LK)光流法,解算图像光流信息,使用惯性导航系统信息辅助光流导航系统,同时也利用光流导航系统的特征点速度信息辅助惯导系统进行导航解算,利用卡尔曼滤波器以融合光流/惯性导航信息得到速度、位置估计信息,并通过仿真实验进行验证.仿真结果表明:该方法能够在纹理丰富和纹理较差的场景下进行精确的速度、位置信息估计,所提出的导航算法符合自主导航的实时性和精确性要求.

Optical Flow/Inertial Integrated Navigation Method for Complex Texture Environment

In order to solve the problem of realization of UAV autonomous navigation in the environment of weak GPS signal or rejection, an optical flow/inertial integrated navigation method suitable for complex texture environment is proposed. Through the improved lucas and kanade (LK) optical flow method based on low-texture scenes, the image optical flow information is calculated, the inertial navigation system information is used to assist the optical flow navigation system, and the characteristic point speed information of the optical flow navigation system is also used to assist the inertial navigation system for navigation solution, use Kalman filter to fuse optical flow/inertial navigation information to obtain speed and position estimation information, and verify it through simulation experiments. Simulation results show that this method can accurately estimate speed and position information in scenes with rich textures and poor textures. The proposed navigation algorithm meets the real-time and accuracy requirements of autonomous navigation.